Mengya Mia Hu Resume Data Science-Full-Time
Mengya Mia Hu Resume Data Science-Full-Time
Mengya Mia Hu Resume Data Science-Full-Time
com/mengyahuUSTC-PU (609)9330077
http://humymia.weebly.com/ MENGYA (MIA) HU mengyah@princeton.edu
EDUCATION
Princeton University (PU) Princeton, NJ Aug. 2015- Expected Jan. 2021
• Ph.D. in Department of Mechanical and Aerospace Engineering
• Courses: Fundamentals of Machine Learning; Machine Learning (ML) & Pattern Recognition; Statistical Analysis
of Financial Data; High Frequency Markets: Models & Data Analysis; Mathematical Methods for Eng. Analysis
University of Science & Technology of China (USTC) Anhui, China Aug. 2011 – July 2015
• B.Eng. (Valedictorian) in Department of Thermal Science & Energy Engineering
• Courses: Computer Programming; Data Structures & Database; Computational Methods; Probability & Statistics.
RESEARCH EXPERIENCE
Quantitative Brokers LLC Quantitative Research Intern May 2019-Aug.2019
Consensus of Multiple Signals for Price Change Prediction
• Created an algorithm combing k-means clustering and linear regression for fitting multivariate relationship
between the signals and the future price change, a crucial part of the trading algorithm
• Achieved faster and better convergence to global optimum for large data sets (millions of rows) by implementing
revised simulated annealing instead of using the existing packages; Data manipulation used packages like dplyr.
• Created data visualization such as specialized 3D plots/videos, based on the basic graphics packages such as
plotly, which helped investigate the convergence of SA and the performance of signal consensus
• All algorithms were implemented in R. Version control was maintained using Github
system, which was essential for validating the feasibility of the space missions and investigating the error budgets
• Developed image processing methods, such as generalized likelihood ratio test (GLRT), to extract exoplanet
images no specialized image processing methods had been developed before. It can help use the observation time
more efficiently as it generates false alarm rates online after each PC image
• Algorithms were in Matlab and run via Batch scheduler on clusters with Linux. Version control used Bitbucket.
SKILLS
Python, R, Matlab; Regression, Classification, Support Vector Machine, Principle Component Analysis, Hypothesis
Testing, Maximum Likelihood Estimation, Kalman Filter, LASSO, ML with Kernels, K-nearest Neighbors, K-
means Clustering , Time Series Analysis (MA, AR, ARMA, ARIMA, ARCH, GARCH), Data Visualization, A/B
Testing, Feature Engineering, Neural Networks(NN), Monte Carlo Simulation, Convolutional Neural Network,
Bayesian Statistics, Cross-Validation.
LEADERSHIP & ACTIVITIES
• Vice President of Princeton University Graduate Society of Women Engineer (Mar. 2018 – May 2019): Co-
initiated this graduate student chapter; Coordinated various social and professional development activities
• Leader of The Debate Team of School of Engineering Science, USTC (Sept. 2011-June 2013): Won 2nd Place
in USTC’s Championships; Recruited and Trained new members; Organized Debate Championships
• Data Science for All, Correlation One (Oct. 2019): Passed online test and selected as 80 out of 900+ to
participate in the two-day training taught by Prof. John Paisley from Columbia University; Proposed, finished and
presented a project analyzing 2017 US commercial airline flight delay patterns with ( ~1,000,000 rows*20
columns data set) using python packages such as pandas, pingouin, matplotlib and seaborn.
AWARDS
• Guo Moruo Scholarship (the highest honor for USTC undergrads, 32 students awarded that year) 2014
SELECTED PUBLICATIONS
Hu, M. M., Sun, H., & Kasdin, N. J. (2019, September). Sequential generalized likelihood ratio test for planet
detection with photon-counting mode. In Techniques and Instrumentation for Detection of Exoplanets IX (Vol.
11117, p. 111171K). International Society for Optics and Photonics.